EP4252182A1 - Method for automatically searching for at least one textile pattern in a composite material reinforcement - Google Patents

Method for automatically searching for at least one textile pattern in a composite material reinforcement

Info

Publication number
EP4252182A1
EP4252182A1 EP21823641.2A EP21823641A EP4252182A1 EP 4252182 A1 EP4252182 A1 EP 4252182A1 EP 21823641 A EP21823641 A EP 21823641A EP 4252182 A1 EP4252182 A1 EP 4252182A1
Authority
EP
European Patent Office
Prior art keywords
textile
reinforcement
composite material
pattern
dimensional image
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
EP21823641.2A
Other languages
German (de)
French (fr)
Inventor
Yanneck WIELHORSKI
Teddy FIXY
Julien Paul SCHNEIDER
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Safran Aircraft Engines SAS
Original Assignee
Safran Aircraft Engines SAS
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Safran Aircraft Engines SAS filed Critical Safran Aircraft Engines SAS
Publication of EP4252182A1 publication Critical patent/EP4252182A1/en
Pending legal-status Critical Current

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/64—Three-dimensional objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
    • G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/01—Arrangements or apparatus for facilitating the optical investigation
    • G01N2021/0181—Memory or computer-assisted visual determination
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N2021/8472—Investigation of composite materials
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N21/00—Investigating or analysing materials by the use of optical means, i.e. using sub-millimetre waves, infrared, visible or ultraviolet light
    • G01N21/84—Systems specially adapted for particular applications
    • G01N21/88—Investigating the presence of flaws or contamination
    • G01N21/8851—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges
    • G01N2021/8883—Scan or image signal processing specially adapted therefor, e.g. for scan signal adjustment, for detecting different kinds of defects, for compensating for structures, markings, edges involving the calculation of gauges, generating models
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/12—Acquisition of 3D measurements of objects
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V2201/00—Indexing scheme relating to image or video recognition or understanding
    • G06V2201/12—Acquisition of 3D measurements of objects
    • G06V2201/122—Computational image acquisition in electron microscopy

Definitions

  • TITLE Automatic search process for at least one textile pattern in a composite material reinforcement
  • the technical field of the invention is that of composite materials and more particularly that of methods for the automatic search of textile patterns for reinforcing composite materials.
  • the present invention relates to a method for searching for at least one textile pattern in a composite material reinforcement and in particular an automatic method for searching for at least one textile pattern in a composite material reinforcement.
  • the present invention also relates to a method for reconstituting the textile geometry of a composite material reinforcement, a method for checking the textile geometry of a composite material reinforcement, a computer program product and a recording medium. making it possible to implement the research method and/or the reconstitution method and/or the control method.
  • a composite material is an assembly comprising at least one textile framework called reinforcement and a binder called matrix.
  • the manufacture of a part in a composite material therefore requires a first step of making the reinforcement, for example by weaving, then a second step of assembly with the matrix, for example by injection.
  • the reinforcement of a part is made in such a way that the textile geometry of the reinforcement conforms to a theoretical textile reinforcement geometry allowing the part to have the desired thermo-physical and/or thermo-mechanical properties.
  • the textile geometry of the reinforcement conforms to a theoretical textile reinforcement geometry allowing the part to have the desired thermo-physical and/or thermo-mechanical properties.
  • These deviations, called textile defects often result in a variation of the thermo-physical and/or thermo-mechanical properties of the part compared to what was planned, and therefore lead to the systematic rejection of the part, considered as defective.
  • the invention offers a solution to the problems mentioned above, by making it possible to precisely detect any textile defects in the reinforcement of a part.
  • a first aspect of the invention relates to a method for automatically searching for at least one given textile pattern in a composite material reinforcement comprising a plurality of textile patterns, each textile pattern comprising a plurality of reinforcing threads arranged in a textile topology, the method comprising the following steps:
  • the artificial neural network makes it possible to automatically detect each textile pattern encountered during its training phase, present in the three-dimensional image acquired.
  • the artificial neural network makes it possible to automatically detect the occurrences of this textile defect in the reinforcement of a composite material.
  • the textile pattern is a textile pattern present in the theoretical textile geometry
  • the artificial neural network makes it possible to detect the occurrences of this textile pattern in the reinforcement of a composite material.
  • the position of each occurrence of the textile pattern with the corresponding position in the theoretical textile geometry, it is possible to verify the conformity of the textile geometry of the reinforcement with the corresponding theoretical textile geometry and to identify possible textile faults and their positions. This comparison can be performed manually or automatically.
  • the research method according to the invention therefore makes it possible to detect any textile defects precisely with respect to a visual inspection method.
  • the method according to the first aspect of the invention may have one or more additional characteristics among the following, considered individually or according to all technically possible combinations.
  • the artificial neural network is a multi-layer perceptron or a convolutional artificial neural network.
  • the artificial neural network is trained in a supervised manner and the training database comprises, for each composite training material, a plurality of composite materials of training, a three-dimensional image of the reinforcement of the composite training material, and for each textile pattern to be detected, the textile topology of the textile pattern and the location of the textile pattern in the three-dimensional image.
  • the artificial neural network is trained to detect, in a three-dimensional image, the textile topology associated with each textile pattern to be detected.
  • the textile topology of each textile pattern to be detected is obtained manually, using mathematical morphology algorithms, an artificial neural network or dedicated software.
  • the three-dimensional image is acquired by X-ray tomography or by transmission electron microscope.
  • a second aspect of the invention relates to a method for automatically reconstituting the textile geometry of a composite material reinforcement comprising a plurality of textile patterns, comprising the steps of the search method according to the first aspect of the invention for each textile pattern of the composite material reinforcement.
  • a third aspect of the invention relates to a method for automatically checking the textile geometry of a composite material reinforcement, comprising the steps of the reconstitution method according to the second aspect of the invention to obtain a reconstitution of the geometry textile of the composite material reinforcement and a comparison step between the reconstitution of the textile geometry of the composite material reinforcement and a theoretical textile geometry.
  • a fourth aspect of the invention relates to a computer configured to implement the steps of the search method according to the first aspect of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
  • a fifth aspect of the invention relates to a computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the steps of the search method according to the first aspect. of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
  • a sixth aspect of the invention relates to a computer-readable recording medium comprising instructions which, when executed by a computer, lead the latter to implement the steps of the method of research according to the first aspect of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
  • Figure 1 shows a three-dimensional image of a composite material reinforcement.
  • Figure 2 shows a digital reconstruction of the architecture of a composite material reinforcement on which a textile pattern is identified.
  • Figure 3 shows a schematic representation of a reinforcing thread on which the skeleton of the reinforcing thread is identified.
  • Figure 4 is a block diagram illustrating the sequence of steps of a search method according to the invention.
  • Figure 5 is a block diagram illustrating the sequence of steps of a reconstitution method according to the invention.
  • Figure 6 is a block diagram illustrating the sequence of steps of a control method according to the invention.
  • a first aspect of the invention relates to an automatic search method for at least one textile pattern in a composite material reinforcement.
  • the reinforcement of a composite material is a textile framework comprising a plurality of reinforcing threads, also called rovings or strands.
  • the reinforcing threads are arranged along at least one axis, called the reinforcing axis.
  • the reinforcement is for example a superposition of reinforcing plies, or reinforcing layers, each comprising a plurality of reinforcing threads.
  • Figure 1 shows a three-dimensional image 301 of the reinforcement 300 of a composite material.
  • the upper reinforcing ply 3001 consists of reinforcing threads 3002 arranged along a Y axis and reinforcing threads 3002 arranged along an X axis.
  • the reinforcing axes X and Y are substantially perpendicular but the reinforcing plies 3001 may include reinforcing threads 3002 arranged along reinforcing axes forming an angle other than 90°.
  • the reinforcement axes X, Y can form an angle of 45°.
  • the reinforcing threads 3002 are arranged together so as to form a particular textile geometry comprising a plurality of textile patterns.
  • Figure 2 shows a digital reconstruction of the architecture of the reinforcement 300 of a composite material on which are identified two occurrences of a textile pattern 3021 presenting a textile topology 302.
  • the term "textile pattern of a reinforcement” means a geometric arrangement of a plurality of reinforcing threads that can be repeated in the reinforcement.
  • the same textile pattern 3021 can therefore have several occurrences in the reinforcement 300 of a composite material.
  • the textile pattern 3021 comprises eight reinforcing threads 3002, two reinforcing threads 3002-1, 3002-2 shown in black arranged along the reinforcing axis Y, three reinforcing threads 3002-3, 3002-4, 3002-5 shown in white and arranged along the reinforcing axis Y and three reinforcing threads 3002-6, 3002-7, 3002-8 shown in gray and arranged along the reinforcing axis X.
  • the thread reinforcement 3002-3 is superimposed on the reinforcement wire 3002-1 and the reinforcement wire 3002-4 is superimposed on the reinforcement wire 3002-2 along the Z axis.
  • 3002-6, 3002-8 are below reinforcing threads 3002-2, 3002-3, 3002-4, 3002-5 and above reinforcing thread 3002-1 and the reinforcement 3002-7 is above reinforcement threads 3002-1, 3002-2, 3002-3, 3002-4, 3002-5.
  • the architecture of the reinforcement 300 of composite material includes two occurrences of the textile pattern 3021 identified by dotted lines.
  • each reinforcing thread 3002 is arranged according to a textile topology 302 corresponding to the skeleton of each reinforcing thread 3002 of the textile pattern 3021.
  • Figure 3 shows a schematic representation of a reinforcing thread 3002 having a skeleton 3022.
  • the skeleton 3022 of a reinforcing thread 3002 arranged along a reinforcing axis X comprises a set of points including for each section 3023 of the reinforcing thread 3002 along a plane 3025 perpendicular to the reinforcing axis X of a plurality of sections 3023 of the reinforcing thread 3002 along a plane 3025 perpendicular to the reinforcing axis X, a point corresponding to the barycenter 3024 of the section 3023 of the reinforcing thread 3002, each plane 3025 being parallel and not confused with other 3025 plans.
  • the skeleton 3022 of a reinforcing thread 3002 can be discrete or continuous. In the latter case, the skeleton 3022 corresponds to an interpolation passing through all the points of the set of points.
  • a textile pattern 3021 can for example be manufactured by braiding, sewing, z-pinning (or z-pinning in English), or even tufting (or tufting in English).
  • a textile pattern 3021 can correspond to a particular textile defect.
  • FIG. 4 is a block diagram illustrating the sequence of steps of the search method 100 according to the invention.
  • a first step 101 of the method 100 consists in acquiring a three-dimensional image 301 of the reinforcement 300 of the composite material.
  • the three-dimensional image 301 is for example acquired by X-ray tomography or by transmission electron microscope with a resolution for example between 1 to 400 ⁇ m, preferably 10 to 200 ⁇ m.
  • the three-dimensional image 301 of the reinforcement 300 of a composite material is acquired by X-ray tomography.
  • a second step 102 of the method 100 consists in using an artificial neural network trained on a training database, to detect each textile pattern 3021 to be searched for present in the three-dimensional image 301 acquired in the first step. 101.
  • An artificial neural network comprises at least one layer of artificial neurons each comprising at least one artificial neuron.
  • the artificial neurons of the artificial neural network are interconnected by synapses and each synapse is assigned a synaptic coefficient.
  • the artificial neural network is for example a multi-layer perceptron or a convolutional artificial neural network, such as the U-net artificial neural network, and in particular the U-net 2D or U-net artificial neural networks. clean 3D.
  • the training makes it possible to train the artificial neural network for a predefined task, by updating the synaptic coefficients so as to minimize the error between the output data provided by the artificial neural network and the real data.
  • output i.e. what the artificial neural network should output to fulfill the predefined task on a certain input datum.
  • the training of the artificial neural network is for example supervised.
  • the training database includes input data, each associated with a real output data.
  • the function of the artificial neural network is to detect each textile pattern 3021 to be detected present in the three-dimensional image 301 previously acquired.
  • the training database therefore includes three-dimensional images 301 of a plurality of composite training materials, as well as data on the textile topology 302 and the position of each textile pattern 3021 to be detected in each image. three-dimensional 301 from the training database.
  • the data includes, for example, the three-dimensional coordinates of each point of the set of points included in the skeleton 3022 of each reinforcing thread 3002 of the textile pattern 3021 .
  • the textile topology 302 of a textile pattern 3021 is for example obtained manually, using mathematical morphology algorithms, an artificial neural network or dedicated software, such as textile modeling, such as TexGen, WiseTex, or Multifil, or random textile geometry generation software.
  • the drive composite material or materials may be identical to or different from the composite material in the reinforcement 300 of which it is desired to search for at least one textile pattern 3021 of the reinforcement 300.
  • the search method 100 it is possible to detect each occurrence of at least one textile pattern 3021 in the reinforcement 300 of a composite material before assembly with the matrix or after assembly with the matrix of the composite material.
  • the three-dimensional images 301 of the training database can therefore be three-dimensional images 301 of reinforcements 300 of composite materials before assembly with their matrices and/or three-dimensional images dimensions 301 of reinforcements 300 of composite materials after assembly with their matrices.
  • a second aspect of the invention relates to a process for automatically reconstituting the textile geometry of the reinforcement 300 of composite material.
  • Figure 5 is a block diagram illustrating the sequence of steps of the reconstitution method 200 according to the invention.
  • reconstruction of the textile geometry of the reinforcement of a composite material means obtaining a digital model of the architecture of the reinforcement of the composite material in which each textile pattern has been identified.
  • the method 200 of reconstitution according to the invention comprises the steps 101, 102 of the method 100 of research according to the invention for each textile pattern 3021 of the reinforcement 300 of composite material.
  • the method 200 comprises three times the steps 101, 102 of the method 100.
  • a third aspect of the invention relates to a process for automatic control of the textile geometry of the reinforcement 300 of composite material.
  • FIG. 6 is a block diagram illustrating the sequence of steps of the control method 400 according to the invention.
  • the control method 400 according to the invention comprises the steps of the reconstitution method 200 according to the invention making it possible to obtain a reconstitution of the textile geometry of the reinforcement 300 of composite material.
  • the control method 400 then comprises a step 401 of comparison between the reconstitution of the textile geometry of the reinforcement 300 of composite material obtained previously and a theoretical textile geometry.
  • “Theoretical textile geometry” means the model or pattern on the basis of which the reinforcement 300 of composite material is made and to which the reinforcement 300 of composite material must conform.
  • the step 401 of comparison between the reconstitution of the textile geometry of the reinforcement 300 of composite material obtained previously and the theoretical textile geometry therefore makes it possible to test the compliance of the reinforcement 300 of composite material and to detect any textile defects.
  • the research method 100, the reconstitution method 200 and the control method 400 are automatic, that is to say they are implemented by a computer.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computing Systems (AREA)
  • Artificial Intelligence (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Software Systems (AREA)
  • Medical Informatics (AREA)
  • Databases & Information Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Molecular Biology (AREA)
  • Mathematical Physics (AREA)
  • General Engineering & Computer Science (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Biomedical Technology (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Data Mining & Analysis (AREA)
  • Treatment Of Fiber Materials (AREA)
  • Image Processing (AREA)
  • Image Analysis (AREA)

Abstract

One aspect of the invention relates to a method for automatically searching for at least one given textile pattern (3021) in a composite material reinforcement (300) comprising a plurality of textile patterns (3021), each textile pattern (3021) comprising a plurality of reinforcing yarns (3002) arranged according to a textile topology (302), the method (100) comprising the following steps: - acquiring a three-dimensional image of the composite material reinforcement (300); - searching for the given textile pattern (3021) in the acquired three-dimensional image, using an artificial neural network trained on a training database to detect the given textile pattern in a three-dimensional image of a composite material reinforcement.

Description

DESCRIPTION DESCRIPTION
TITRE : Procédé de recherche automatique d’au moins un motif textile dans un renfort de matériau compositeTITLE: Automatic search process for at least one textile pattern in a composite material reinforcement
DOMAINE TECHNIQUE DE L’INVENTION TECHNICAL FIELD OF THE INVENTION
[0001] Le domaine technique de l’invention est celui des matériaux composites et plus particulièrement celui des procédés de recherche automatique de motifs textiles de renfort de matériaux composites. [0001] The technical field of the invention is that of composite materials and more particularly that of methods for the automatic search of textile patterns for reinforcing composite materials.
[0002] La présente invention concerne un procédé de recherche d’au moins un motif textile dans un renfort de matériau composite et en particulier un procédé automatique de recherche d’au moins un motif textile dans un renfort de matériau composite. La présente invention concerne également un procédé de reconstitution de la géométrie textile d’un renfort de matériau composite, un procédé de contrôle de la géométrie textile d’un renfort de matériau composite, un produit-programme d’ordinateur et un support d’enregistrement permettant de mettre en œuvre le procédé de recherche et/ou le procédé de reconstitution et/ou le procédé de contrôle. The present invention relates to a method for searching for at least one textile pattern in a composite material reinforcement and in particular an automatic method for searching for at least one textile pattern in a composite material reinforcement. The present invention also relates to a method for reconstituting the textile geometry of a composite material reinforcement, a method for checking the textile geometry of a composite material reinforcement, a computer program product and a recording medium. making it possible to implement the research method and/or the reconstitution method and/or the control method.
ARRIERE-PLAN TECHNOLOGIQUE DE L’INVENTION TECHNOLOGICAL BACKGROUND OF THE INVENTION
[0003] Un matériau composite est un assemblage comportant au moins une ossature textile appelée renfort et un liant appelé matrice. La fabrication d’une pièce dans un matériau composite nécessite donc une première étape de confection du renfort, par exemple par tissage, puis une deuxième étape d’assemblage avec la matrice, par exemple par injection. [0003] A composite material is an assembly comprising at least one textile framework called reinforcement and a binder called matrix. The manufacture of a part in a composite material therefore requires a first step of making the reinforcement, for example by weaving, then a second step of assembly with the matrix, for example by injection.
[0004] Le renfort d’une pièce est confectionné de manière que la géométrie textile du renfort soit conforme à une géométrie textile de renfort théorique permettant à la pièce d’avoir les propriétés thermo-physiques et/ou thermomécaniques souhaitées. Cependant, il est courant que des écarts soient constatés entre la géométrie de renfort réelle de la pièce et la géométrie de renfort théorique. Ces écarts, appelés défauts textiles, ont souvent pour conséquence une variation des propriétés thermo-physiques et/ou thermomécaniques de la pièce par rapport à ce qui a été prévu, et entraînent donc le rejet systématique de la pièce, considérée comme défectueuse. [0004] The reinforcement of a part is made in such a way that the textile geometry of the reinforcement conforms to a theoretical textile reinforcement geometry allowing the part to have the desired thermo-physical and/or thermo-mechanical properties. However, it is common for discrepancies to be seen between the actual reinforcement geometry of the part and the theoretical reinforcement geometry. These deviations, called textile defects, often result in a variation of the thermo-physical and/or thermo-mechanical properties of the part compared to what was planned, and therefore lead to the systematic rejection of the part, considered as defective.
[0005] Actuellement, l’identification des défauts textiles est réalisée par contrôle visuel, ce qui rend l’opération extrêmement chronophage, une simple aube LEAP™ comportant plusieurs milliers de fils de carbone à contrôler, et sujette à de nombreuses erreurs, car il est parfois impossible, même à l’œil nu, de distinguer des défauts textiles. [0005] Currently, the identification of textile defects is carried out by visual inspection, which makes the operation extremely time-consuming, a simple LEAP™ blade comprising several thousand carbon threads to be checked, and subject to numerous errors, because it is sometimes impossible, even with the naked eye, to distinguish textile defects.
[0006] Il existe donc un besoin de pouvoir contrôler la conformité de la géométrie textile du renfort d’une pièce avec la géométrie textile théorique, avec un risque réduit d’erreurs. [0006] There is therefore a need to be able to check the conformity of the textile geometry of the reinforcement of a part with the theoretical textile geometry, with a reduced risk of errors.
RESUME DE L’INVENTION SUMMARY OF THE INVENTION
[0007] L’invention offre une solution aux problèmes évoqués précédemment, en permettant de détecter de manière précise, les éventuels défauts textiles du renfort d’une pièce. The invention offers a solution to the problems mentioned above, by making it possible to precisely detect any textile defects in the reinforcement of a part.
[0008] Un premier aspect de l’invention concerne un procédé de recherche automatique d’au moins un motif textile donné dans un renfort de matériau composite comportant une pluralité de motifs textiles, chaque motif textile comportant une pluralité de fils de renfort agencés selon une topologie textile, le procédé comportant les étapes suivantes : A first aspect of the invention relates to a method for automatically searching for at least one given textile pattern in a composite material reinforcement comprising a plurality of textile patterns, each textile pattern comprising a plurality of reinforcing threads arranged in a textile topology, the method comprising the following steps:
Acquisition d’une image tri-dimensionnelle du renfort de matériau composite ; Acquisition of a three-dimensional image of the composite material reinforcement;
Recherche, à l’aide d’un réseau de neurones artificiels entraîné sur une base de données d’entraînement pour détecter le motif textile donné dans une image tri-dimensionnelle de renfort de matériau composite, du motif textile donné dans l’image tri-dimensionnelle acquise. Search, using an artificial neural network trained on a training database to detect the textile pattern given in a three-dimensional image of reinforcement of composite material, of the textile pattern given in the three-dimensional image. dimension acquired.
[0009] Grâce à l’invention, le réseau de neurones artificiels permet de détecter automatiquement chaque motif textile rencontré pendant sa phase d’entraînement, présent dans l’image tri-dimensionnelle acquise. Thanks to the invention, the artificial neural network makes it possible to automatically detect each textile pattern encountered during its training phase, present in the three-dimensional image acquired.
[0010] Si le motif textile est un défaut textile connu, le réseau de neurones artificiels permet de détecter automatiquement les occurrences de ce défaut textile dans le renfort d’un matériau composite. [0010] If the textile pattern is a known textile defect, the artificial neural network makes it possible to automatically detect the occurrences of this textile defect in the reinforcement of a composite material.
[0011] Si le motif textile est un motif textile présent dans la géométrie textile théorique, le réseau de neurones artificiels permet de détecter les occurrences de ce motif textile dans le renfort d’un matériau composite. Ainsi, en comparant la position de chaque occurrence du motif textile avec la position correspondante dans la géométrie textile théorique, il est possible de vérifier la conformité de la géométrie textile du renfort avec la géométrie textile théorique correspondante et d’identifier d’éventuels défauts textiles et leurs positions. Cette comparaison peut être réalisée manuellement ou automatiquement. [0011] If the textile pattern is a textile pattern present in the theoretical textile geometry, the artificial neural network makes it possible to detect the occurrences of this textile pattern in the reinforcement of a composite material. Thus, by comparing the position of each occurrence of the textile pattern with the corresponding position in the theoretical textile geometry, it is possible to verify the conformity of the textile geometry of the reinforcement with the corresponding theoretical textile geometry and to identify possible textile faults and their positions. This comparison can be performed manually or automatically.
[0012] Le procédé de recherche selon l’invention permet donc de détecter les éventuels défauts textiles de manière précise par rapport à une méthode de contrôle visuel. The research method according to the invention therefore makes it possible to detect any textile defects precisely with respect to a visual inspection method.
[0013] Outre les caractéristiques qui viennent d’être évoquées dans le paragraphe précédent, le procédé selon le premier aspect de l’invention peut présenter une ou plusieurs caractéristiques complémentaires parmi les suivantes, considérées individuellement ou selon toutes les combinaisons techniquement possibles. [0013] In addition to the characteristics which have just been mentioned in the previous paragraph, the method according to the first aspect of the invention may have one or more additional characteristics among the following, considered individually or according to all technically possible combinations.
[0014] Selon une variante de réalisation, le réseau de neurones artificiels est un perceptron multi-couches ou un réseau de neurones artificiels convolutif. [0014] According to a variant embodiment, the artificial neural network is a multi-layer perceptron or a convolutional artificial neural network.
[0015] Selon une variante de réalisation compatible avec la variante de réalisation précédente, le réseau de neurones artificiels est entraîné de manière supervisée et la base de données d’entraînement comporte pour chaque matériau composite d’entraînement d’une pluralité de matériaux composites d’entraînement, une image tri dimensionnelle du renfort du matériau composite d’entraînement, et pour chaque motif textile à détecter, la topologie textile du motif textile et la localisation du motif textile dans l’image tri-dimensionnelle. [0015] According to a variant embodiment compatible with the preceding variant embodiment, the artificial neural network is trained in a supervised manner and the training database comprises, for each composite training material, a plurality of composite materials of training, a three-dimensional image of the reinforcement of the composite training material, and for each textile pattern to be detected, the textile topology of the textile pattern and the location of the textile pattern in the three-dimensional image.
[0016] Ainsi, le réseau de neurones artificiels est entraîné à détecter dans une image tri-dimensionnelle, la topologie textile associée à chaque motif textile à détecter. [0016] Thus, the artificial neural network is trained to detect, in a three-dimensional image, the textile topology associated with each textile pattern to be detected.
[0017] Selon une sous-variante de réalisation de la variante de réalisation précédente, la topologie textile de chaque motif textile à détecter est obtenue manuellement, à l’aide d’algorithmes de morphologie mathématique, d’un réseau de neurones artificiels ou d’un logiciel dédié. According to a sub-variant embodiment of the previous variant embodiment, the textile topology of each textile pattern to be detected is obtained manually, using mathematical morphology algorithms, an artificial neural network or dedicated software.
[0018] Selon une variante de réalisation compatible avec les variantes de réalisation précédentes, l’image tri-dimensionnelle est acquise par tomographie par rayons X ou par microscope électronique en transmission. [0019] Un deuxième aspect de l’invention concerne un procédé de reconstitution automatique de la géométrie textile d’un renfort de matériau composite comportant une pluralité de motifs textiles, comportant les étapes du procédé de recherche selon le premier aspect de l’invention pour chaque motif textile du renfort de matériau composite. According to a variant embodiment compatible with the preceding variants, the three-dimensional image is acquired by X-ray tomography or by transmission electron microscope. A second aspect of the invention relates to a method for automatically reconstituting the textile geometry of a composite material reinforcement comprising a plurality of textile patterns, comprising the steps of the search method according to the first aspect of the invention for each textile pattern of the composite material reinforcement.
[0020] Ainsi, il est possible d’obtenir automatiquement une reconstitution complète de l’architecture du renfort du matériau composite dans laquelle chaque point du renfort a été associé à un motif textile connu. [0020] Thus, it is possible to automatically obtain a complete reconstitution of the architecture of the reinforcement of the composite material in which each point of the reinforcement has been associated with a known textile pattern.
[0021] Un troisième aspect de l’invention concerne un procédé de contrôle automatique de la géométrie textile d’un renfort de matériau composite, comportant les étapes du procédé de reconstitution selon le deuxième aspect de l’invention pour obtenir une reconstitution de la géométrie textile du renfort de matériau composite et une étape de comparaison entre la reconstitution de la géométrie textile du renfort de matériau composite et une géométrie textile théorique. A third aspect of the invention relates to a method for automatically checking the textile geometry of a composite material reinforcement, comprising the steps of the reconstitution method according to the second aspect of the invention to obtain a reconstitution of the geometry textile of the composite material reinforcement and a comparison step between the reconstitution of the textile geometry of the composite material reinforcement and a theoretical textile geometry.
[0022] Ainsi, il est possible d’obtenir automatiquement tous les écarts entre la géométrie textile reconstruite et la géométrie textile théorique, et donc les éventuels défauts textiles du renfort du matériau composite. [0022] Thus, it is possible to automatically obtain all the differences between the reconstructed textile geometry and the theoretical textile geometry, and therefore the possible textile defects of the reinforcement of the composite material.
[0023] Un quatrième aspect de l’invention concerne un calculateur configuré pour mettre en œuvre les étapes du procédé de recherche selon le premier aspect de l’invention et/ou du procédé de reconstitution selon le deuxième aspect de l’invention et/ou du procédé de contrôle selon le troisième aspect de l’invention. A fourth aspect of the invention relates to a computer configured to implement the steps of the search method according to the first aspect of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
[0024] Un cinquième aspect de l’invention concerne un produit-programme d’ordinateur comprenant des instructions qui, lorsque le programme est exécuté par un ordinateur, conduisent celui-ci à mettre en œuvre les étapes du procédé de recherche selon le premier aspect de l’invention et/ou du procédé de reconstitution selon le deuxième aspect de l’invention et/ou du procédé de contrôle selon le troisième aspect de l’invention. A fifth aspect of the invention relates to a computer program product comprising instructions which, when the program is executed by a computer, lead the latter to implement the steps of the search method according to the first aspect. of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
[0025] Un sixième aspect de l’invention concerne un support d’enregistrement lisible par ordinateur comprenant des instructions qui, lorsqu’elles sont exécutées par un ordinateur, conduisent celui-ci à mettre en œuvre les étapes du procédé de recherche selon le premier aspect de l’invention et/ou du procédé de reconstitution selon le deuxième aspect de l’invention et/ou du procédé de contrôle selon le troisième aspect de l’invention. A sixth aspect of the invention relates to a computer-readable recording medium comprising instructions which, when executed by a computer, lead the latter to implement the steps of the method of research according to the first aspect of the invention and/or of the reconstitution method according to the second aspect of the invention and/or of the control method according to the third aspect of the invention.
[0026] L’invention et ses différentes applications seront mieux comprises à la lecture de la description qui suit et à l’examen des figures qui l’accompagnent. The invention and its various applications will be better understood on reading the following description and on examining the accompanying figures.
BREVE DESCRIPTION DES FIGURES BRIEF DESCRIPTION OF FIGURES
[0027] Les figures sont présentées à titre indicatif et nullement limitatif de l’invention. The figures are presented for information only and in no way limit the invention.
La figure 1 montre une image tri-dimensionnelle d’un renfort de matériau composite. Figure 1 shows a three-dimensional image of a composite material reinforcement.
La figure 2 montre une reconstitution numérique de l’architecture d’un renfort de matériau composite sur laquelle est identifié un motif textile.Figure 2 shows a digital reconstruction of the architecture of a composite material reinforcement on which a textile pattern is identified.
La figure 3 montre une représentation schématique d’un fil de renfort sur lequel est identifié le squelette du fil de renfort. Figure 3 shows a schematic representation of a reinforcing thread on which the skeleton of the reinforcing thread is identified.
La figure 4 est un schéma synoptique illustrant l’enchaînement des étapes d’un procédé de recherche selon l’invention. Figure 4 is a block diagram illustrating the sequence of steps of a search method according to the invention.
La figure 5 est un schéma synoptique illustrant l’enchaînement des étapes d’un procédé de reconstitution selon l’invention. Figure 5 is a block diagram illustrating the sequence of steps of a reconstitution method according to the invention.
La figure 6 est un schéma synoptique illustrant l’enchaînement des étapes d’un procédé de contrôle selon l’invention. Figure 6 is a block diagram illustrating the sequence of steps of a control method according to the invention.
DESCRIPTION DETAILLEE DETAILED DESCRIPTION
[0028] Sauf précision contraire, un même élément apparaissant sur des figures différentes présente une référence unique. Unless specified otherwise, the same element appearing in different figures has a single reference.
[0029] Un premier aspect de l’invention concerne un procédé de recherche automatique d’au moins un motif textile dans un renfort de matériau composite. A first aspect of the invention relates to an automatic search method for at least one textile pattern in a composite material reinforcement.
[0030] Le renfort d’un matériau composite est une ossature textile comportant une pluralité de fils de renfort, encore appelés mèches ou torons. [0030] The reinforcement of a composite material is a textile framework comprising a plurality of reinforcing threads, also called rovings or strands.
[0031] Les fils de renfort sont agencés selon au moins un axe, appelé axe de renfort. [0032] Le renfort est par exemple une superposition de plis de renfort, ou couches de renfort, comprenant chacun une pluralité de fils de renfort. [0031] The reinforcing threads are arranged along at least one axis, called the reinforcing axis. The reinforcement is for example a superposition of reinforcing plies, or reinforcing layers, each comprising a plurality of reinforcing threads.
[0033] [Fig. 1] La figure 1 montre une image tri-dimensionnelle 301 du renfort 300 d’un matériau composite. [0033] [Fig. 1] Figure 1 shows a three-dimensional image 301 of the reinforcement 300 of a composite material.
[0034] Sur la figure 1 , le pli de renfort 3001 supérieur est constitué de fils de renfort 3002 agencés selon un axe Y et de fils de renfort 3002 agencés selon un axe X. [0035] Sur la figure 1 , les axes de renfort X et Y sont sensiblement perpendiculaires mais les plis de renfort 3001 peuvent comporter des fils de renfort 3002 agencés selon des axes de renfort formant un angle différent de 90°. Par exemple, les axes de renfort X, Y peuvent former un angle de 45°. [0034] In Figure 1, the upper reinforcing ply 3001 consists of reinforcing threads 3002 arranged along a Y axis and reinforcing threads 3002 arranged along an X axis. [0035] In Figure 1, the reinforcing axes X and Y are substantially perpendicular but the reinforcing plies 3001 may include reinforcing threads 3002 arranged along reinforcing axes forming an angle other than 90°. For example, the reinforcement axes X, Y can form an angle of 45°.
[0036] Dans le renfort 300 d’un matériau composite, les fils de renfort 3002 sont agencés entre eux de manière à former une géométrie textile particulière comportant une pluralité de motifs textiles. In the reinforcement 300 of a composite material, the reinforcing threads 3002 are arranged together so as to form a particular textile geometry comprising a plurality of textile patterns.
[0037] [Fig. 2] La figure 2 montre une reconstitution numérique de l’architecture du renfort 300 d’un matériau composite sur laquelle sont identifiées deux occurrences d’un motif textile 3021 présentant une topologie textile 302. [0037] [Fig. 2] Figure 2 shows a digital reconstruction of the architecture of the reinforcement 300 of a composite material on which are identified two occurrences of a textile pattern 3021 presenting a textile topology 302.
[0038] On entend par « motif textile d’un renfort », un agencement géométrique d’une pluralité de fils de renfort pouvant se répéter dans le renfort. [0038] The term "textile pattern of a reinforcement" means a geometric arrangement of a plurality of reinforcing threads that can be repeated in the reinforcement.
[0039] Un même motif textile 3021 peut donc présenter plusieurs occurrences dans le renfort 300 d’un matériau composite. [0039] The same textile pattern 3021 can therefore have several occurrences in the reinforcement 300 of a composite material.
[0040] Sur la figure 2, le motif textile 3021 comporte huit fils de renfort 3002, deux fils de renfort 3002-1 , 3002-2 représentés en noir agencés selon l’axe de renfort Y, trois fils de renfort 3002-3, 3002-4, 3002-5 représentés en blanc et agencés selon l’axe de renfort Y et trois fils de renfort 3002-6, 3002-7, 3002-8 représentés en gris et agencés selon l’axe de renfort X. Le fil de renfort 3002-3 est superposé au fil de renfort 3002-1 et le fil de renfort 3002-4 est superposé au fil de renfort 3002-2 selon l’axe Z. [0041] Selon l’axe Z, les fils de renfort 3002-6, 3002-8 sont en dessous des fils de renfort 3002-2, 3002-3, 3002-4, 3002-5 et au-dessus du fil de renfort 3002-1 et le fil de renfort 3002-7 est au-dessus des fils de renfort 3002-1 , 3002-2, 3002-3, 3002-4, 3002- 5. In Figure 2, the textile pattern 3021 comprises eight reinforcing threads 3002, two reinforcing threads 3002-1, 3002-2 shown in black arranged along the reinforcing axis Y, three reinforcing threads 3002-3, 3002-4, 3002-5 shown in white and arranged along the reinforcing axis Y and three reinforcing threads 3002-6, 3002-7, 3002-8 shown in gray and arranged along the reinforcing axis X. The thread reinforcement 3002-3 is superimposed on the reinforcement wire 3002-1 and the reinforcement wire 3002-4 is superimposed on the reinforcement wire 3002-2 along the Z axis. 3002-6, 3002-8 are below reinforcing threads 3002-2, 3002-3, 3002-4, 3002-5 and above reinforcing thread 3002-1 and the reinforcement 3002-7 is above reinforcement threads 3002-1, 3002-2, 3002-3, 3002-4, 3002-5.
[0042] Sur la figure 2, l’architecture du renfort 300 de matériau composite comporte deux occurrences du motif textile 3021 identifiées par des pointillés. In Figure 2, the architecture of the reinforcement 300 of composite material includes two occurrences of the textile pattern 3021 identified by dotted lines.
[0043] Au sein d’un motif textile 3021 , chaque fil de renfort 3002 est agencé selon une topologie textile 302 correspondant au squelette de chaque fil de renfort 3002 du motif textile 3021 . Within a textile pattern 3021, each reinforcing thread 3002 is arranged according to a textile topology 302 corresponding to the skeleton of each reinforcing thread 3002 of the textile pattern 3021.
[0044] [Fig. 3] La figure 3 montre une représentation schématique d’un fil de renfort 3002 présentant un squelette 3022. [0044] [Fig. 3] Figure 3 shows a schematic representation of a reinforcing thread 3002 having a skeleton 3022.
[0045] Comme représenté sur la figure 3, le squelette 3022 d’un fil de renfort 3002 agencé selon un axe de renfort X comporte un ensemble de points incluant pour chaque section 3023 du fil de renfort 3002 selon un plan 3025 perpendiculaire à l’axe de renfort X d’une pluralité de sections 3023 du fil de renfort 3002 selon un plan 3025 perpendiculaire à l’axe de renfort X, un point correspondant au barycentre 3024 de la section 3023 du fil de renfort 3002, chaque plan 3025 étant parallèle et non confondu avec les autres plans 3025. As shown in Figure 3, the skeleton 3022 of a reinforcing thread 3002 arranged along a reinforcing axis X comprises a set of points including for each section 3023 of the reinforcing thread 3002 along a plane 3025 perpendicular to the reinforcing axis X of a plurality of sections 3023 of the reinforcing thread 3002 along a plane 3025 perpendicular to the reinforcing axis X, a point corresponding to the barycenter 3024 of the section 3023 of the reinforcing thread 3002, each plane 3025 being parallel and not confused with other 3025 plans.
[0046] Le squelette 3022 d’un fil de renfort 3002 peut être discret, ou continu. Dans ce dernier cas, le squelette 3022 correspond à une interpolation passant par tous les points de l’ensemble de points. The skeleton 3022 of a reinforcing thread 3002 can be discrete or continuous. In the latter case, the skeleton 3022 corresponds to an interpolation passing through all the points of the set of points.
[0047] Un motif textile 3021 peut par exemple être fabriqué par tressage, couture, épinglage z (ou z-pinning en anglais), ou encore touffetage (ou tufting en anglais). A textile pattern 3021 can for example be manufactured by braiding, sewing, z-pinning (or z-pinning in English), or even tufting (or tufting in English).
[0048] Un motif textile 3021 peut correspondre à un défaut textile particulier. A textile pattern 3021 can correspond to a particular textile defect.
[0049] Le procédé selon le premier aspect de l’invention permet d’obtenir, pour chaque motif textile 3021 à rechercher, le nombre de motifs textiles 3021 détectés et leur localisation ou position, dans le renfort 300 du matériau composite. [0050] [Fig. 4] La figure 4 est un schéma synoptique illustrant l’enchaînement des étapes du procédé 100 de recherche selon l’invention. The method according to the first aspect of the invention makes it possible to obtain, for each textile pattern 3021 to be searched, the number of textile patterns 3021 detected and their location or position, in the reinforcement 300 of the composite material. [0050] [Fig. 4] FIG. 4 is a block diagram illustrating the sequence of steps of the search method 100 according to the invention.
[0051] Une première étape 101 du procédé 100 consiste à acquérir une image tri dimensionnelle 301 du renfort 300 du matériau composite. A first step 101 of the method 100 consists in acquiring a three-dimensional image 301 of the reinforcement 300 of the composite material.
[0052] L’image tri-dimensionnelle 301 est par exemple acquise par tomographie par rayons X ou par microscope électronique en transmission avec une résolution par exemple comprise entre de 1 à 400 pm de préférence de 10 à 200 pm. The three-dimensional image 301 is for example acquired by X-ray tomography or by transmission electron microscope with a resolution for example between 1 to 400 μm, preferably 10 to 200 μm.
[0053] Sur la figure 4, l’image tri-dimensionnelle 301 du renfort 300 d’un matériau composite est acquise par tomographie par rayons X. In Figure 4, the three-dimensional image 301 of the reinforcement 300 of a composite material is acquired by X-ray tomography.
[0054] Une deuxième étape 102 du procédé 100 consiste à utiliser un réseau de neurones artificiels entraîné sur une base de données d’entraînement, pour détecter chaque motif textile 3021 à rechercher présent dans l’image tri-dimensionnelle 301 acquise à la première étape 101. A second step 102 of the method 100 consists in using an artificial neural network trained on a training database, to detect each textile pattern 3021 to be searched for present in the three-dimensional image 301 acquired in the first step. 101.
[0055] Un réseau de neurones artificiels comporte au moins une couche de neurones artificiels comportant chacune au moins un neurone artificiel. Les neurones artificiels du réseau de neurones artificiels sont reliés entre eux par des synapses et à chaque synapse est affecté un coefficient synaptique. An artificial neural network comprises at least one layer of artificial neurons each comprising at least one artificial neuron. The artificial neurons of the artificial neural network are interconnected by synapses and each synapse is assigned a synaptic coefficient.
[0056] Le réseau de neurones artificiels est par exemple un perceptron multi- couches ou un réseau de neurones artificiels convolutif, tel que le réseau de neurones artificiels U-net, et en particulier les réseaux de neurones artificiels U-net 2D ou U-net 3D. The artificial neural network is for example a multi-layer perceptron or a convolutional artificial neural network, such as the U-net artificial neural network, and in particular the U-net 2D or U-net artificial neural networks. clean 3D.
[0057] L’entraînement permet d’entraîner le réseau de neurones artificiels à une tâche prédéfinie, en mettant à jour les coefficients synaptiques de manière à minimiser l’erreur entre la donnée de sortie fournie par le réseau de neurones artificiels et la vraie donnée de sortie, c’est-à-dire ce que le réseau de neurones artificiels devrait fournir en sortie pour remplir la tâche prédéfinie sur une certaine donnée d’entrée. [0058] L’entraînement du réseau de neurones artificiels est par exemple supervisé. Dans ce cas, la base de données d’entraînement comporte des données d’entrée, chacune associée à une vraie donnée de sortie. The training makes it possible to train the artificial neural network for a predefined task, by updating the synaptic coefficients so as to minimize the error between the output data provided by the artificial neural network and the real data. output i.e. what the artificial neural network should output to fulfill the predefined task on a certain input datum. The training of the artificial neural network is for example supervised. In this case, the training database includes input data, each associated with a real output data.
[0059] Le réseau de neurones artificiels a pour fonction de détecter chaque motif textile 3021 à détecter présent dans l’image tri-dimensionnelle 301 précédemment acquise. The function of the artificial neural network is to detect each textile pattern 3021 to be detected present in the three-dimensional image 301 previously acquired.
[0060] La base de données d’entraînement comporte donc des images tri dimensionnelles 301 d’une pluralité de matériaux composites d’entraînement, ainsi que des données sur la topologie textile 302 et la position de chaque motif textile 3021 à détecter dans chaque image tridimensionnelle 301 de la base de données d’entraînement. The training database therefore includes three-dimensional images 301 of a plurality of composite training materials, as well as data on the textile topology 302 and the position of each textile pattern 3021 to be detected in each image. three-dimensional 301 from the training database.
[0061] Les données comportent par exemple les coordonnées tri-dimensionnelles de chaque point de l’ensemble de points inclus dans le squelette 3022 de chaque fil de renfort 3002 du motif textile 3021 . The data includes, for example, the three-dimensional coordinates of each point of the set of points included in the skeleton 3022 of each reinforcing thread 3002 of the textile pattern 3021 .
[0062] La topologie textile 302 d’un motif textile 3021 est par exemple obtenue manuellement, à l’aide d’algorithmes de morphologie mathématique, d’un réseau de neurones artificiels ou d’un logiciel dédié, tel qu’un logiciel de modélisation textile, comme TexGen, WiseTex, ou Multifil, ou un logiciel de génération aléatoire de géométrie textile. The textile topology 302 of a textile pattern 3021 is for example obtained manually, using mathematical morphology algorithms, an artificial neural network or dedicated software, such as textile modeling, such as TexGen, WiseTex, or Multifil, or random textile geometry generation software.
[0063] Le ou les matériaux composites d’entraînement peuvent être identiques ou différents du matériau composite dans le renfort 300 duquel on veut rechercher au moins un motif textile 3021 du renfort 300. The drive composite material or materials may be identical to or different from the composite material in the reinforcement 300 of which it is desired to search for at least one textile pattern 3021 of the reinforcement 300.
[0064] Grâce au procédé 100 de recherche selon l’invention, il est possible de détecter chaque occurrence d’au moins un motif textile 3021 dans le renfort 300 d’un matériau composite avant assemblage avec la matrice ou après assemblage avec la matrice du matériau composite. Thanks to the search method 100 according to the invention, it is possible to detect each occurrence of at least one textile pattern 3021 in the reinforcement 300 of a composite material before assembly with the matrix or after assembly with the matrix of the composite material.
[0065] Les images tri-dimensionnelles 301 de la base de données d’entraînement peuvent donc être des images tri-dimensionnelles 301 de renforts 300 de matériaux composites avant assemblage avec leurs matrices et/ou des images tri- dimensionnelles 301 de renforts 300 de matériaux composites après assemblage avec leurs matrices. The three-dimensional images 301 of the training database can therefore be three-dimensional images 301 of reinforcements 300 of composite materials before assembly with their matrices and/or three-dimensional images dimensions 301 of reinforcements 300 of composite materials after assembly with their matrices.
[0066] Un deuxième aspect de l’invention concerne un procédé de reconstitution automatique de la géométrie textile du renfort 300 de matériau composite. A second aspect of the invention relates to a process for automatically reconstituting the textile geometry of the reinforcement 300 of composite material.
[0067] [Fig. 5] La figure 5 est un schéma synoptique illustrant l’enchaînement des étapes du procédé 200 de reconstitution selon l’invention. [0067] [Fig. 5] Figure 5 is a block diagram illustrating the sequence of steps of the reconstitution method 200 according to the invention.
[0068] On entend par « reconstitution de la géométrie textile du renfort d’un matériau composite », l’obtention d’un modèle numérique de l’architecture du renfort du matériau composite dans laquelle chaque motif textile a été identifié. The term "reconstruction of the textile geometry of the reinforcement of a composite material" means obtaining a digital model of the architecture of the reinforcement of the composite material in which each textile pattern has been identified.
[0069] Le procédé 200 de reconstitution selon l’invention comporte les étapes 101 , 102 du procédé 100 de recherche selon l’invention pour chaque motif textile 3021 du renfort 300 de matériau composite. The method 200 of reconstitution according to the invention comprises the steps 101, 102 of the method 100 of research according to the invention for each textile pattern 3021 of the reinforcement 300 of composite material.
[0070] Sur la figure 5, le procédé 200 comporte trois fois les étapes 101 , 102 du procédé 100. In Figure 5, the method 200 comprises three times the steps 101, 102 of the method 100.
[0071] Un troisième aspect de l’invention concerne un procédé de contrôle automatique de la géométrie textile du renfort 300 de matériau composite. A third aspect of the invention relates to a process for automatic control of the textile geometry of the reinforcement 300 of composite material.
[0072] [Fig. 6] La figure 6 est un schéma synoptique illustrant l’enchaînement des étapes du procédé 400 de contrôle selon l’invention. [0072] [Fig. 6] Figure 6 is a block diagram illustrating the sequence of steps of the control method 400 according to the invention.
[0073] Le procédé 400 de contrôle selon l’invention comporte les étapes du procédé 200 de reconstitution selon l’invention permettant d’obtenir une reconstitution de la géométrie textile du renfort 300 de matériau composite. The control method 400 according to the invention comprises the steps of the reconstitution method 200 according to the invention making it possible to obtain a reconstitution of the textile geometry of the reinforcement 300 of composite material.
[0074] Le procédé 400 de contrôle selon l’invention comporte ensuite une étape 401 de comparaison entre la reconstitution de la géométrie textile du renfort 300 de matériau composite obtenue précédemment et une géométrie textile théorique. [0075] Par « géométrie textile théorique », on entend le modèle ou patron sur la base duquel le renfort 300 de matériau composite est confectionné et auquel le renfort 300 de matériau composite doit être conforme. [0076] L’étape 401 de comparaison entre la reconstitution de la géométrie textile du renfort 300 de matériau composite obtenue précédemment et la géométrie textile théorique permet donc de tester la conformité du renfort 300 de matériau composite et de détecter d’éventuels défauts textiles. [0077] Le procédé 100 de recherche, le procédé 200 de reconstitution et le procédé 400 de contrôle sont automatiques, c’est-à-dire qu’ils sont mis en œuvre par un calculateur. The control method 400 according to the invention then comprises a step 401 of comparison between the reconstitution of the textile geometry of the reinforcement 300 of composite material obtained previously and a theoretical textile geometry. “Theoretical textile geometry” means the model or pattern on the basis of which the reinforcement 300 of composite material is made and to which the reinforcement 300 of composite material must conform. The step 401 of comparison between the reconstitution of the textile geometry of the reinforcement 300 of composite material obtained previously and the theoretical textile geometry therefore makes it possible to test the compliance of the reinforcement 300 of composite material and to detect any textile defects. The research method 100, the reconstitution method 200 and the control method 400 are automatic, that is to say they are implemented by a computer.

Claims

REVENDICATIONS
[Revendication 1] Procédé (100) de recherche automatique d’au moins un motif textile (3021) donné dans un renfort (300) de matériau composite comportant une pluralité de motifs textiles (3021), chaque motif textile (3021) comportant une pluralité de fils de renfort (3002) agencés selon une topologie textile (302), le procédé (100) étant caractérisé en ce qu’il comporte les étapes suivantes : [Claim 1] Method (100) of automatic search for at least one textile pattern (3021) given in a reinforcement (300) of composite material comprising a plurality of textile patterns (3021), each textile pattern (3021) comprising a plurality reinforcing threads (3002) arranged according to a textile topology (302), the method (100) being characterized in that it comprises the following steps:
- Acquisition d’une image tri-dimensionnelle (301) du renfort (300) de matériau composite (101) ; - Acquisition of a three-dimensional image (301) of the reinforcement (300) of composite material (101);
- Recherche, à l’aide d’un réseau de neurones artificiels entraîné sur une base de données d’entraînement pour détecter le motif textile donné dans une image tri-dimensionnelle de renfort de matériau composite, du motif textile (3021) donné dans l’image tri-dimensionnelle (301) acquise (102). - Search, using an artificial neural network trained on a training database to detect the textile pattern given in a three-dimensional image of reinforcement of composite material, of the textile pattern (3021) given in the three-dimensional image (301) acquired (102).
[Revendication 2] Procédé (100) selon la revendication 1 , caractérisé en ce que le réseau de neurones artificiels est un perceptron multi-couches ou un réseau de neurones artificiels convolutif. [Claim 2] Method (100) according to claim 1, characterized in that the artificial neural network is a multi-layer perceptron or a convolutional artificial neural network.
[Revendication s] Procédé (100) selon l’une quelconque des revendications précédentes, caractérisé en ce que le réseau de neurones artificiels est entraîné de manière supervisée et la base de données d’entraînement comporte pour chaque matériau composite d’entraînement d’une pluralité de matériaux composites d’entraînement, une image tri-dimensionnelle (301) du renfort (300) du matériau composite d’entraînement, et pour chaque motif textile (3021) à détecter, la topologie textile (302) du motif textile (3021 ) et la localisation du motif textile (3021 ) dans l’image tri-dimensionnelle (301). [Claim s] Method (100) according to any one of the preceding claims, characterized in that the artificial neural network is trained in a supervised manner and the training database comprises, for each composite training material, a plurality of composite training materials, a three-dimensional image (301) of the reinforcement (300) of the composite training material, and for each textile pattern (3021) to be detected, the textile topology (302) of the textile pattern (3021 ) and the location of the textile pattern (3021) in the three-dimensional image (301).
[Revendication 4] Procédé (100) selon la revendication 3, caractérisé en ce que la topologie textile (302) de chaque motif textile (3021) à détecter est obtenue manuellement, à l’aide d’algorithmes de morphologie mathématique, d’un réseau de neurones artificiels ou d’un logiciel dédié. [Claim 4] Method (100) according to Claim 3, characterized in that the textile topology (302) of each textile pattern (3021) to be detected is obtained manually, using mathematical morphology algorithms, from a artificial neural network or dedicated software.
[Revendication s] Procédé (100) selon l’une quelconque des revendications précédentes, caractérisé en ce que l’image tri-dimensionnelle (301) est acquise par tomographie par rayons X ou par microscope électronique en transmission. [Claim s] Method (100) according to any one of the preceding claims, characterized in that the three-dimensional image (301) is acquired by X-ray tomography or by transmission electron microscope.
[Revendication 6] Procédé (200) de reconstitution automatique de la géométrie textile d’un renfort (300) de matériau composite comportant une pluralité de motifs textiles (3021), caractérisé en ce qu’il comporte les étapes du procédé (100) de recherche selon l’une quelconque des revendications précédentes pour chaque motif textile (3021) du renfort (300) de matériau composite, pour obtenir le nombre de détections de chaque motif textile (3021) et la localisation de chaque détection dans le renfort (300) de matériau composite. [Claim 6] Method (200) for automatically reconstituting the textile geometry of a reinforcement (300) of composite material comprising a plurality of textile patterns (3021), characterized in that it comprises the steps of the method (100) of search according to any one of the preceding claims for each textile pattern (3021) of the reinforcement (300) of composite material, to obtain the number of detections of each textile pattern (3021) and the location of each detection in the reinforcement (300) of composite material.
[Revendication 7] Procédé (400) de contrôle automatique de la géométrie textile d’un renfort (300) de matériau composite, caractérisé en ce qu’il comporte les étapes du procédé (200) de reconstitution selon la revendication 6 pour obtenir une reconstitution de la géométrie textile du renfort (300) de matériau composite et une étape (401) de comparaison entre la reconstitution de la géométrie textile du renfort (300) de matériau composite et une géométrie textile théorique. [Claim 7] Process (400) for automatic control of the textile geometry of a reinforcement (300) of composite material, characterized in that it comprises the steps of the process (200) for reconstitution according to Claim 6 for obtaining a reconstitution of the textile geometry of the reinforcement (300) of composite material and a step (401) of comparison between the reconstitution of the textile geometry of the reinforcement (300) of composite material and a theoretical textile geometry.
[Revendication 8] Calculateur configuré pour mettre en œuvre les étapes du procédé (100) de recherche selon l’une quelconque des revendications 1 à 5 et/ou du procédé (200) de reconstitution selon la revendication 6 et/ou du procédé (400) de contrôle selon la revendication 7. [Claim 8] Computer configured to implement the steps of the search method (100) according to any one of Claims 1 to 5 and/or of the reconstitution method (200) according to Claim 6 and/or of the method (400) ) control according to claim 7.
[Revendication 9] Produit-programme d’ordinateur comprenant des instructions qui, lorsque le programme est exécuté par un ordinateur, conduisent celui-ci à mettre en œuvre les étapes du procédé (100) de recherche selon l’une quelconque des revendications 1 à 5 et/ou du procédé (200) de reconstitution selon la revendication 6 et/ou du procédé (400) de contrôle selon la revendication 7. [Claim 9] Computer program product comprising instructions which, when the program is executed by a computer, cause the latter to implement the steps of the search method (100) according to any one of Claims 1 to 5 and/or the reconstitution method (200) according to claim 6 and/or the control method (400) according to claim 7.
[Revendication 10] Support d’enregistrement lisible par ordinateur comprenant des instructions qui, lorsqu’elles sont exécutées par un ordinateur, conduisent celui-ci à mettre en œuvre les étapes du procédé (100) de recherche selon l’une quelconque des revendications 1 à 5 et/ou du procédé (200) de reconstitution selon la revendication 6 et/ou du procédé (400) de contrôle selon la revendication 7. [Claim 10] A computer-readable recording medium comprising instructions which, when executed by a computer, cause the computer to perform the steps of the search method (100) according to any of claims 1 to 5 and/or the reconstitution method (200) according to claim 6 and/or the control method (400) according to claim 7.
EP21823641.2A 2020-11-30 2021-11-22 Method for automatically searching for at least one textile pattern in a composite material reinforcement Pending EP4252182A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
FR2012365A FR3116928B1 (en) 2020-11-30 2020-11-30 Method for automatically searching for at least one textile pattern in a composite material reinforcement
PCT/FR2021/052056 WO2022112697A1 (en) 2020-11-30 2021-11-22 Method for automatically searching for at least one textile pattern in a composite material reinforcement

Publications (1)

Publication Number Publication Date
EP4252182A1 true EP4252182A1 (en) 2023-10-04

Family

ID=75108429

Family Applications (1)

Application Number Title Priority Date Filing Date
EP21823641.2A Pending EP4252182A1 (en) 2020-11-30 2021-11-22 Method for automatically searching for at least one textile pattern in a composite material reinforcement

Country Status (5)

Country Link
US (1) US20240005679A1 (en)
EP (1) EP4252182A1 (en)
CN (1) CN116569025A (en)
FR (1) FR3116928B1 (en)
WO (1) WO2022112697A1 (en)

Also Published As

Publication number Publication date
WO2022112697A1 (en) 2022-06-02
CN116569025A (en) 2023-08-08
FR3116928B1 (en) 2023-09-22
FR3116928A1 (en) 2022-06-03
US20240005679A1 (en) 2024-01-04

Similar Documents

Publication Publication Date Title
EP2989449B1 (en) Machine for weaving or winding a fibrous texture that allows checks for anamoly to be made using image analysis
WO2004038617A2 (en) Method and device for synthesising an electrical architecture
EP3000003A1 (en) Method for defining fibre paths from a stress grid or stress curves
CA2765241A1 (en) Method and device for image processing, particularly for medical image processing
Kim et al. Deep learning based automatic defect classification in through-silicon Via process: FA: Factory automation
FR3050826A1 (en) METHOD OF NON-DESTRUCTIVE CONTROL BY LEARNING
FR3050274A1 (en) METHOD OF TOPOLOGICAL ANALYSIS OF A COMPOSITE PIECE
EP1327156B1 (en) Method for isolating an integrated circuit element
EP4252182A1 (en) Method for automatically searching for at least one textile pattern in a composite material reinforcement
EP2145175B1 (en) Method and machine for inspecting transparent or translucent articles
WO2023006627A1 (en) Method for automatically detecting defects in the components of a circuit board
EP3000002B1 (en) Method for defining fibre paths from a transfer surface
Kähler et al. Anomaly detection for industrial surface inspection: application in maintenance of aircraft components
EP4073751B1 (en) Method for automatically reconstituting the reinforcing architecture of a composite material
EP3809366A1 (en) Manufacturing method
Adolphy et al. Method for automated structuring of product data and its applications
WO2011042386A1 (en) Method and device for the automatic inspection of a cable spool
FR3122273A1 (en) Method for automatically generating an artificial image of the reinforcement of a composite material
WO2014135770A1 (en) Method for managing data relative to motor vehicles with a view to the subsequent graphic generation of electrical diagrams of electrical systems
WO2022029385A1 (en) Method of characterising a woven fibrous structure
FR2976559A1 (en) Technical method for maintaining rotary wing aircraft, involves establishing maintenance list including incriminated member cited in elected location and/or provisional rule, and allowing operator to act technically on member
WO2022223902A1 (en) Method for automatically reconstituting the architecture of the reinforcement of a composite material
FR3100331A1 (en) TEST METHOD AND SYSTEM FOR EVALUATING THE QUALITY OF A MICRO-OPTICAL PROJECTION AND / OR MULTI-CHANNEL SUB-WAVELENGTH OPTICAL PROJECTION UNIT
FR3126802A1 (en) Method for characterizing a weaving anomaly detected in a three-dimensional image of a part made from a woven composite material
FR3105872A1 (en) COMPARISON PROCESS OF AT LEAST TWO FLIGHT PLANS AMONG A FIRST FLIGHT PLAN AND A SECOND FLIGHT PLAN, COMPUTER PROGRAM PRODUCT AND ASSOCIATED COMPARISON MODULE

Legal Events

Date Code Title Description
STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: UNKNOWN

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE INTERNATIONAL PUBLICATION HAS BEEN MADE

PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: REQUEST FOR EXAMINATION WAS MADE

17P Request for examination filed

Effective date: 20230628

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

DAV Request for validation of the european patent (deleted)
DAX Request for extension of the european patent (deleted)